bad interaction of rassociative and tellsimp



On 3/27/11, Henry Baker <hbaker1 at pipeline.com> wrote:

> Was the problem something trivial, or was it something more fundamental?

It was pretty trivial -- the code to handle rassociative assumed
that the operator of an expression after simplification is the same
as it was before. That's not true for the zand/zor rule.

> I'm going to be nosing around quite a bit using tellsimp & tellsimpafter, so
> I may find a number of additional bugs/misfeatures of this type.

OK. My advice is to spend only a little time trying to figure out
strange behavior -- chances are good that there are some
known limitations that you will run into -- so feel free to post a
message on the mailing list when you run into problems.

best

Robert Dodier